There’s a lot of noise right now around as ChatGPT started testing ads the day after Anthropic mocked the idea of those very ads in a series of Super Bowl spots. Unsurprisingly, brands are asking the question: should we be testing this, or not yet?
The honest answer isn’t a simple yes or no. It depends on how this product actually comes to market not just how it’s being talked about.
Here’s how we’re thinking about it.
What we know so far
Early details suggest ChatGPT ads will enter the market at a premium. Initial CPMs are expected to start around $60, with minimum spends reportedly in the $200,000 range and reporting limited to impressions and clicks.
That price point immediately stands out. Premium CTV inventory on platforms like Hulu, Disney+, and Netflix typically sits closer to the $30–35 CPM range, while Meta is often well below that. Even Google Search can effectively exceed $60 CPMs in competitive categories, but advertisers don’t tend to think about it that way because it’s measured in CPCs and tied directly to conversion intent.
So the price point is undeniably premium without the measurement maturity advertisers are used to.
Ads will be clearly labeled and ChatGPT is adamant that ads do not influence the answers ChatGPT will provide. Advertisers will not see personal details or conversations with ChatGPT. While OpenAI initially stated they would not “sell user data to advertisers”, they appear to mean that in the literal sense of not profiting directly of the sale of data. They will allow targeting based off of that data: what you prompt and discuss, as well as past chats.
The Bear Case
The bear case for ChatGPT ads is straightforward: price and maturity are misaligned.
Charging some of the highest CPMs on the web without clear conversion visibility is a bold starting position. The significant reporting limitations open up questions on product maturity. While there’s talk of advertisers already participating in beta programs, there’s limited visibility into who those advertisers are, how scalable activation is, or how quickly brands can actually get live. This rollout feels rushed, and in some ways a response to growing momentum for competitors like Anthropic’s Claude and Google’s Gemini.
Targeting is another open question. If early ChatGPT ads rely on broad, topic-level targeting rather than prompt- or keyword-level intent, much of the platform’s theoretical advantage is reduced. Reaching users inside an LLM only matters if advertisers can align closely with context and intent, not just general categories. Higher conversion rates, and the higher CPMs those conversion rates justify, will come from being able to show an ad to someone looking for trail running shoes that come in size 14 for runners with poor arch support, not everyone lumped into a topic for athletic wear.
The Bull Case
The bull case is that ChatGPT ads could represent a meaningful evolution of intent-based advertising.
Online advertising changed fundamentally when Google enabled brands to reach users actively expressing intent at the point in time they were expressing it. If ChatGPT eventually supports granular prompt-level targeting, it could take that model further by combining intent with unusually rich context.
There are strong reasons to believe this could matter. ChatGPT’s audience skews toward highly engaged, high-value users. Early data suggests that traffic from large language models already converts at higher than average rates. And today, understanding prompt-level demand is extremely difficult via 3P measurement partners — even directional impression data could help brands better size the true AI search opportunity in their category.
There’s also a strategic argument for early learning. ChatGPT has hired experienced leaders from companies like Meta and Instacart, and history suggests that once ad products are live, they tend to evolve quickly. Brands that engage early — thoughtfully —often learn faster than those that wait for perfection.
So, should your brand advertise on ChatGPT?
Right now, the decision comes down to risk tolerance and objectives.
If targeting remains broad, budgets are tightly ROI-constrained, or clear conversion measurement is non-negotiable, it likely makes sense to wait. If, however, targeting becomes more granular, your brand is in growth mode, or you’re trying to understand the size and shape of AI-driven demand, a deliberate test could be worthwhile.
The key word there is deliberate.
Bottom line
ChatGPT ads aren’t a must-buy today. But they also should not be dismissed as just hype.
Most brands would find more immediate value by focusing on their organic AI visibility strategy, where impact is already being felt across multiple platforms without paid minimums.
This is likely the beginning of a new advertising surface, not a finished channel. The brands that benefit most won’t be the ones that rush in blindly — they’ll be the ones that test with intent, understand the trade-offs, and stay close as the product evolves.
For now, the smartest posture is informed curiosity: neither hype-driven nor dismissive.
